Binary tree-structured statistical classification algorithms and properties of 56 model alkyl nucleophiles were brought to bear on two problems of experimental pharmacology and toxicology. Each rat of a learning sample of 745 was administered one compound and autopsied to determine the presence of duodenal ulcer or adrenal hemorrhagic necrosis. The cited statistical classification schemes were then applied to these outcomes and 67 features of the compounds to ascertain those characteristics that are associated with biologic activity. For predicting duodenal ulceration, dipole moment, melting point, and solubility in octanol are particularly important, while for predicting adrenal necrosis, important features include the number of sulfhydryl groups and double bonds. These methods may constitute inexpensive but powerful ways to screen untested compounds for possible organ-specific toxicity. Mechanisms for the etiology and pathogenesis of the duodenal and adrenal lesions are suggested, as are additional avenues for drug design.Duodenal ulcer disease is a frequent and poorly understood disorder. Quantitative structure-activity studies of chemical duodenal ulcerogens are of great interest, since they can yield toxicologic information regarding the etiology and mechanisms of ulceration, suggest pharmacologic approaches, and ultimately bear upon clinical practice. Duodenal ulcers are 2-4 times more frequent than gastric ulcers in humans, yet much more is known about the pathophysiology of gastric mucosal lesions than about duodenal ulceration (1). This is due in part to the lack of availability of requisite animal models of duodenal ulcer disease.It has long been known that gastric ulcers and erosions can be induced easily in both humans and laboratory animals by a variety of chemicals (in toxic doses) that elicit stress response. Selye (2) was the first to recognize that during stress a common set of pathogenetic mechanisms link many of the different etiologic factors. Until recently, animal models of duodenal ulcers were scarce (1). Available methods, such as pantothenic acid deficiency in certain strains of rats, the prolonged infusion of secretagogues, or the local application of acetic acid on the duodenum, were cumbersome and typically produced a low yield of duodenal ulcers (1, 3, 4). The first rapidly developing experimental duodenal ulcer was produced by subcutaneous administration of propionitrile, which consistently induced solitary, and often perforating, duodenal ulcers in 80%o of rats (5). Subsequently, cysteamine was also found to produce duodenal ulcers in rats (6). This model offered advantages since the ulcers developed more rapidly and consistently than after propionitrile administration, and only one to three doses were required. Since then, several other compounds have been shown to be duodenal ulcerogens, including 3,4-toluenediamine (7), 3,4-toluenedithiol, and n-butyronitrile (8, 9). The latter two chemicals were discovered during our initial structure-activity studies, which revealed th...